An Electrochemical Cell Configuration Incorporating an Ion

Int. J. Electrochem. Sci., 4 (2009) 1 - 8

International Journal of

ELECTROCHEMICAL

SCIENCE www.electrochemsci.org

An Electrochemical Cell Configuration Incorporating an Ion

Conducting Membrane Separator between Reference and

Working Electrode

K.J.J. Mayrhofer*, Sean J. Ashton, J. Kreuzer and M. Arenz

Institut für Physikalische Chemie, Technische Universität München, Garching D-85748 *E-mail: [email protected]

Received: 8 August 2008 / Accepted: 18 November 2008 / Published: 20 December 2008

Ionic impurities can significantly interfere with certain electrochemical measurements and should

therefore be completely avoided. However, standard reference electrodes such as the widely used

Saturated Calomel Electrode (SCE) and the Ag/AgCl electrode contain saturated chloride solutions.

Typically, this chloride solution is only separated from the electrochemical cell electrolyte by a porous

glass frit, which still allows chlorides to diffuse through and contaminate the electrochemical system.

In order to avoid this undesired effect, a new electrochemical cell setup is presented here. The

modification incorporates an ion conducting membrane that serves to further separate chloride

containing reference electrodes from the electrochemical cell, and thereby actively prevents the diffusion of chlorides to the surface of the working electrode. The benefits of this modification, in

particular for long-term electrochemical measurements in which the electrodes are positioned closely, is demonstrated by means of the investigation of the oxygen reduction reaction (ORR) on

polycrystalline Pt in both a single- and three-compartment electrochemical cell.

Keywords: electrocatalysis, reference electrodes, chloride contamination, long-term measurements,

Tschurl modification

1. INTRODUCTION

Techniques involving electrochemical cells have a long standing tradition in the study of the

thermodynamic and kinetic behavior of half cell reactions, and are currently widely used in

electrocatalysis, corrosion and sensor research [1, 2]. Depending on the application, various cell

designs and experimental assemblies are employed [3]. For electrochemical investigations setups with

three electrodes are preferred over those involving two electrodes due to the reduction in

Int. J. Electrochem. Sci., Vol. 4, 2009

2

uncompensated resistance [4, 5]. Moreover, the electrodes can either be situated in one single

electrochemical cell compartment, or in several, separate electrode compartments connected via

electrolytic bridges. The latter configuration has the advantage that interaction between reactions at

each of the electrodes is avoided, which is crucial for studies that are prone to impurities [6].

In a three-electrode, three-compartment cell the counter electrode is shielded to prevent excessive

diffusion of gaseous products into the working electrode compartment. If no undesirable products are

produced at the counter electrode, a simple microporous frit is sufficient for this purpose. The effective

separation of the reference electrode, on the other hand, can be decisive for the quality and accuracy of

electrochemical measurements; in particular for reference electrodes, which contain ions that can

influence certain reactions at the working electrode. The most convenient and therefore most widely

used reference electrodes, namely the Saturated Calomel Electrode (SCE) or the Ag/AgCl electrode,

are embedded into saturated KCl solution in order to maintain their highly constant reference potential

[7]. These electrodes are normally contacted to the electrolyte via a porous frit that serves as a salt

bridge. The large chloride concentration gradient across the porous frit, however, leads to the diffusion

of chloride ions into the electrochemical cell electrolyte, which can severely impact adsorption and

reaction kinetics at the surface of the working electrode [8, 9]. These gradual changes in chloride

concentration are in general hard to detect and can often be overlooked during the course of an

electrochemical measurement. Consequently, experimental data can easily be misinterpreted, and

electrocatalytic activities can be erroneous.

In order to avoid contamination from chlorides, we have modified the setup of our three-

compartment cell by introducing an ion conducting membrane (Nafion®) in order to prevent the

chlorides in the reference electrode from diffusing to the working electrode. Nafion® consists of a

polytetrafluoroethylene skeletal structure with sulfate groups located at the end of protruding side

chains. The latter structure introduces partial hydrophilic character to the otherwise hydrophobic

polymer backbone. Consequently solvation shells form around the sulfate groups inside the membrane

and water molecules can be absorbed. Accordingly, channels allowing the diffusion of water and small

positive ions such as H+ or Na+ are established inside the membrane. Negative ions are however,

repelled by the negative charge of the sulfate groups and ideally cannot enter these channels and

diffuse through the membrane [10]. In this study we present the application of the new setup for the

measurement of oxygen reduction on platinum, a reaction that is particularly sensitive to chloride

impurities [8, 11].

2. EXPERIMENTAL PART

The electrochemical measurements were conducted in either a Teflon [12, 13] or a standard glass

three-compartment electrochemical cell, using a rotating-disc electrode (RDE) setup, potentiostat

(Princeton Applied Research) and rotation controller (Radiometer Analytical). A saturated calomel

reference electrode (SCE, Schott), Pt mesh counter electrode and a polycrystalline bulk Platinum

sample (5 mm diameter, 0.196 cm2 geometrical surface area) working electrode were used in each

measurement of this study. The electrolytes were prepared using Millipore® water and suprapure


3

HClO4 (Merck). The polycrystalline sample was cleaned with concentrated perchloric acid for

approximately 5 minutes and then rinsed thoroughly with Milipore® water prior to each experiment.

For the oxygen reduction reaction (ORR), cyclovoltammograms in argon and the hydrogen

evolution/oxidation reaction (HER/HOR), the electrolyte was purged and saturated with the respective

gases (Air Liquide). The exact potential of the SCE versus the reversible hydrogen electrode (RHE)

was measured shortly after each ORR measurement. All measurements were conducted at room

temperature.

Figure 1. Teflon Cell setup and magnification of the modified reference electrode side compartment

(A) applied in this study. WE = working electrode; RE = reference electrode; CE = counter electrode.

Figure 1 shows a schematic of the Teflon cell and the modified reference electrode side

compartment. The working electrode (WE) is located at the tip of the rotating disc electrode assembly

(RDE), and is submerged in the electrolyte of the main compartment. The counter electrode (CE) is

connected to the main compartment by a small hole in the side wall. The incoming gases meanwhile

are purged through a porous Teflon plug into the electrolyte of the main compartment. The reference

electrode (RE) is simply free-standing inside the ‘Tschurl Modification’ (A) in the RE compartment.

The Haber-Luggin capillary, which connects the RE compartment with the main compartment and

ends close to the surface of the working electrode, is necessary for precise potential control and the

avoidance of any uncompensated resistance in the measurements. The ‘Tschurl Modification’ itself (A)

consists of a Teflon body-tubing and screw fitting with a hole in the base. The ion conducting

membrane (Nafion®) is pressed tightly, with the aid of a Teflon spacer, to the bottom of the screw

fitting in order to prevent the leakage of electrolyte and efficiently separate the electrolyte in the

‘Tschurl Modification’ from the electrolyte in the reference electrode compartment. The three feet at

the base ensure that diffusion to the membrane is not physically blocked at any time.


4

3. RESULTS AND DISCUSSION

In order to evaluate the impact of chlorides on the ORR on polycrystalline Pt, KCl was added to

the oxygen purged electrolyte in concentrations ranging from 10-7

mol L-1

to 2x10-5

mol L-1

. The

polarization curves closely resemble previously published curves of the ORR [6, 14], upon which the

effect of chlorides on several Pt systems has been described previously [6, 8, 11]. For the quantitative

analysis of the ORR activity in this study, the cyclovoltammograms (CVs) are corrected for the non-

faradaic background by subtracting the CVs recorded in Ar-purged electrolyte. No correction for the

uncompensated resistance is applied. Since the kinetic region of the ORR may shift by up to 200 mV

towards more negative potentials in the course of an experiment due to the impact of chlorides, a

comparison of kinetic currents at a certain, fixed potential is unfeasible [ 14]. Therefore the apparent

activity for the ORR is specified by the potential at half of the diffusion limited current (ORR half-

wave potential) of the anodic potential scan at 1600 rpm throughout this study. The dependence of the

ORR half-wave potentials on the chloride concentration is depicted in Figure 2.

0.8

0.7

VR

HE

6

10-7

2 3 4 5 6

10-6

2 3 4 5 6

10-5

2 3

Cl- conc. /M

ORR half-wave potential0.1 M HClO4;1600 rpm; 293 K; 50 mV/s

Figure 2. Impact of Chlorides on the ORR on polycrystalline Pt in 0.1 mol L

-1 HClO4 electrolyte

(Teflon cell).

At a chloride concentration of 10-7 mol L-1 the ORR activity on polycrystalline Pt deviates only by

1 mV from the half-wave potential in a chloride-free electrolyte (855 mV), which is within the

experimental error of the measurement. Above a concentration of 5x10-7

mol L-1

, however, the

decrease in the activity (6 mV shift in the ORR half-wave potential) becomes more significant. This

would already be equivalent to a loss in kinetic current density of approximately 15%. With increasing

chloride concentration the ORR curves shift further negative, at 2x10-5 mol L-1 (highest concentration

used in this study) the half-wave potential is as low as 711 mV, a decrease of 144 mV. This clearly

demonstrates that in order to determine the precise electrocatalytic activities of ORR catalysts, even

minor contamination by chlorides must be avoided.


5

3.1 Reference electrode – side compartment

In order to relate these findings to practical measurements, the influence of the Saturated Calomel

Electrode (SCE) on the ORR measurement was investigated using a number of experimental

configurations. The results obtained from long-term measurements in a standard glass cell and a Teflon

cell with and without the use of our modified reference electrode side compartment are presented in

Figure 3.

0.9

0.8

0.7

0.6

OR

R H

alf

Wa

ve

Po

ten

tia

l /V

RH

E

3002001000

time /min

standard glass cell

Teflon cell without Tschurl mod.

Teflon cell with Tschurl mod.

side compartment

Figure 3. Long-term stability of the measurement of the ORR on polycrystalline Pt in 0.1 mol L-1

HClO4 electrolyte in a standard glass and a Teflon electrochemical cell setup. The SCE was placed in the side compartment in all cases. After 290 minutes, convection of electrolyte from the side

compartment to the main compartment was enforced in both the glass cell (orange), and the Teflon cell with (brown) and without (red) the modified reference electrode side compartment.

The SCE was placed in the RE compartment of the three electrode setup described in Figure 1. The

activity for the ORR did not change over 290 min, within the experimental error of 2 mV. This

indicates that measurements can be conducted over a period of 5 hours and the system remains free of

contaminations. A slightly reduced activity is observed for each measurement in the glass cell;

however, this is attributed to a slightly different cell geometry and the resulting difference in the

uncompensated resistance in comparison to the Teflon cell. In the standard glass cell diffusion of

chlorides from the reference electrode side compartment to the main compartment is additionally

inhibited by a stop-cock, and a very thin Luggin capillary. In the Teflon cell the distance from the

reference compartment to the main compartment also seems to be sufficient, so that in this time period

chlorides do not diffuse to the working electrode surface. This limited effect of chlorides can be

attributed to the deliberate design of the three-compartment cells. Therefore, in this cell geometry the

modified reference electrode side compartment does not enhance the stability of ORR measurements to

any significant degree.


6

If, however, convection of the electrolyte from the side compartment to the main compartment

occurs, as it is often the case in the course of an electrocatalytic measurement, then the apparent

activity for the ORR reduces drastically. This can be seen in Figure 3, where after the initial 5 hours of

measurement we forced convection between the different compartments. Evidently, chlorides had

continuously diffused through the porous frit of the reference electrode and accumulated in the

reference side compartment of the glass and the Teflon cell. When the chloride containing electrolyte

in the side compartment is mixed with the electrolyte in the main compartment, these chlorides can

diffuse to the surface of the working electrode. At potentials above 0.5 VRHE they adsorb strongly onto

the Pt surface and as a consequence block active sites for the ORR. The effect is almost identical in the

glass cell and the Teflon cell without modified reference electrode side compartment, suggesting that

in both cases an equal amount of chlorides had diffused from the reference electrode into the cell

electrolyte. According to the decrease in activity observed in Figure 2, the concentration is estimated to

be greater than 2x10-5

mol L-1

. If the reference electrode is placed inside the modified reference

electrode side compartment, however, the diffusion of chlorides into the electrolyte can be reduced

significantly. After forced convection the activity drops only by 14 mV, which infers a final chloride

concentration of approximately 2x10-6 mol L-1 that can be ascribed to the minor co-ion diffusion

through the Nafion membrane [15].

3.2 Reference electrode – main compartment

In simple, commercially available electrochemical cells the reference electrode is often placed

inside the main compartment, and not in a separate side compartment. The influence of the ‘Tschurl

Modification’ in such a setup was therefore additionally studied.

0.9

0.8

0.7

0.6

0.5

0.4

0.3

Ha

lf W

ave

Po

ten

tial

/VR

HE

3002001000

time /min

3 membranes

1 membrane

no membrane

Figure 4. Long-term stability of the measurement of the ORR on polycrystalline Pt in 0.1 mol L

-1

HClO4 electrolyte in a Teflon Cell with the Calomel reference electrode placed in the main

compartment.


7

Figure 4 clearly demonstrates the serious effect of chloride contamination of the electrolyte on the

ORR activity, when no modified reference electrode side compartment is used. During the first ORR

measurement, after only approximately 10 minutes the half-wave potential had already shifted

negatively by about 150 mV, despite thoroughly rinsing of the reference electrode with Millipore water

beforehand. Note that part of this lower initial activity observed in Figure 3 compared to Figure 2, is

due to the difference in uncompensated resistance. Over the course of the experiment chlorides

continuously diffuse through the porous frit, a process observed in the corresponding decrease in ORR

activity. If, however, the reference electrode is placed into the ‘Tschurl Modification’ with only one

Nafion membrane, no chloride contamination is initially evident and the deterioration in the ORR

activity over time is significantly reduced. A superior chloride diffusion inhibiting effect can be

achieved by employing 3 membranes in the reference electrode compartment. In this case the ORR

activity remains almost unaltered after 290 minutes, with a negative shift of the half-wave potential of

only 10 mV. However, it is clear that the Nafion membranes do not completely prevent the diffusion of

chlorides into the cell electrolyte. There is the possibility that other types of ion conducting membranes

with different polymer chain length, equivalent weight and thickness can be used in order to optimize

the reference electrode compartment [10].

Chlorides can significantly interfere with electrochemical measurements; even at concentrations as

low as 10-7

mol L-1

already decisive deviations in activity occur for the ORR on Pt. Standard reference

electrodes such as the SCE or the Ag/AgCl electrode, which include saturated solutions of chlorides,

can consequently act as a possible source of systematic errors. In order to avoid this chloride

contamination, a newly developed reference electrode compartment is presented. An ion conducting

membrane separates the electrolyte surrounding the reference electrode from the working electrode

compartment, and thereby inhibits the diffusion of chlorides toward the electrolyte in the main

compartment. Additional investigation is, however, necessary in order to optimize the membrane to

further inhibit chloride diffusion, in particular for enabling long-term experiments for reactions that are

sensitive to chloride impurities.

4. CONCLUSIONS

Chlorides can significantly interfere with electrochemical measurements; even at concentrations

as low as 10-7 mol L-1 already decisive deviations in activity occur for the ORR on Pt. Standard

reference electrodes such as the SCE or the Ag/AgCl electrode, which include saturated solutions of

chlorides, can consequently act as a possible source of systematic errors. In order to avoid this chloride

contamination, a newly developed reference electrode compartment is presented. An ion conducting

membrane separates the electrolyte surrounding the reference electrode from the working electrode

compartment, and thereby inhibits the diffusion of chlorides toward the electrolyte in the main

compartment. Additional investigation is, however, necessary in order to optimize the membrane to

further inhibit chloride diffusion, in particular for enabling long-term experiments for reactions that are

sensitive to chloride impurities.


8

ACKNOWLEDGEMENTS

This work was supported by the DFG through the Emmy-Noether project ARE852/1-1. Karl J.J.

Mayrhofer thanks the Austrian FWF, which supports him with an Erwin-Schrödinger Scholarship.

Furthermore we would like to thank the machine shop at the TU Munich for producing the Teflon cell

and the Tschurl modification. We thank M. Tschurl for fruitful discussions.

References

1. A. J. Bard, L. R. Faulkner, Electrochemical Methods. John Wiley & Sons, Inc.: 2000.

2. W. Vielstich, C. H. Hamann, Elektrochemie. 4 ed.; Wiley-VCH: 2005; p 672. 3. R. E. White, Electrochemical Cell Design. Springer: 1984; p 400.

4. J. C. Myland, K. B. Oldham, Analalytical Chemistry, 72 (17)(2000) 3972. 5. W. Oelssner, F. Berthold, U. Guth, Materials and Corrosion-Werkstoffe Und Korrosion

57 (6) (2006) 455.

6. N. M. Markovic, P. N. Ross, Surface Science Reports 45 (4-6) (2002) 117.

7. D. J. G. Ives, G. J. Janz, Reference Electrodes, Theory and Practice. Academic Press: 1961; p

651.

8. V. Stamenkovic, N. M. Markovic, P. N. Ross, Journal of Electroanalytical Chemistry,

500 (1-2) (2001) 44.

9. Z. Xinsheng, S. Gongquan, J. Luhua, C. Weimin, T. Shuihua, Z. Bing, X. Qin,

Electrochemical and Solid-State Letters, 8 (3) (2005) A149.

10. N. P. Berezina, N. A. Kononenko, O.A. Dyomina, N. P. Gnusin, Advances in Colloid and

Interface Science, 139 (1-2) (2008) 3.

11. T. J. Schmidt, U. A. Paulus, H. A. Gasteiger, R. J. Behm, Journal of Electroanalytical

Chemistry, 508 (1-2) (2001) 41.

12. K. J. J. Mayrhofer, G. K. H. Wiberg, M. Arenz, Journal of The Electrochemical Society

155, (1) (2008) P1.

13. K. J. J. Mayrhofer, A. S. Crampton, G. K. H. Wiberg, M. Arenz, Journal of The

Electrochemical Society, 155 (6) (2008) P78.

14. K. J. J. Mayrhofer, D. Strmcnik, B. B. Blizanac, V. Stamenkovic, M. Arenz, N. M. Markovic, Electrochimica Acta, 53 (2008) 3181.

15. A. Herrera, H. L. Yeager, Journal of The Electrochemical Society, 134, (10) (1987) 2446.

© 2009 by ESG (www.electrochemsci.org)

Documents

An Electrochemical Cell Configuration Incorporating an Ion